Stroke occurs in 7-8% of pediatric patients with Sickle Cell Disease (Hb SS) and is a major cause of morbidity. The 46-90% rate of recurrence can be reduced to less than 10% by chronic blood transfusion. Secondary prevention programs based on transfusion have become the standard of care, but most patients have already suffered irreversible brain injury. Prevention of first stroke is now possible using transcranial Doppler (TCD) ultrasound which can detect arterial abnormalities which place these children at high risk for cerebral infarction. Prediction is based on the detection of high arterial blood flow rates which are associated with arterial narrowing and subsequent cerebral infarction. Combining predictive screening and a potentially effective treatment, prevention of stroke before brain injury occurs is now possible. A randomized, controlled trial is needed to determine if first stroke in high risk children is significantly reduced by prophylactic transfusion. This randomized, controlled, multicenter trial of efficacy of periodic blood transfusion will test the hypothesis that reduction of sickle hemoglobin to 30% or less with transfusion will reduce first time stroke by at least 75% compared to supportive medical care. Primary endpoints will be clinically evident symptoms of cerebral infarction with consistent findings on Magnetic Resonance Imaging (MRI), symptomatic intracranial hemorrhage and death from any cause; secondary endpoints will be asymptomatic brain lesions, detected by MRI in brain areas not implicated in primary endpoints. Hematologic characteristics of the high risk group will be analyzed and serum and DNA samples frozen for future analyses. Study design calls for 6-month start-up, TCD screening and randomization conducted during months 6-24, and observation for stroke from entry through month 54. TCD will be interpreted blindly. Endpoints will be adjudicated and MRIs interpreted centrally by blinded panels. Sample size calculation is based on prospective data relating TCD velocity to risk of stroke and a randomized criterion associated with a 46% risk of cerebral infarction over 39 months of observation. Sample size (60 each group) is sufficient to detect 75% reduction in the primary endpoints of stroke at 90% power allowing for up to a 10% dropout. The pool of asymptomatic children to be screened at the 12 centers participating in this is two times that needed to find the 120 patients. (See companion proposal for Data Coordinating Center: Donald Brambla, Ph.D., P.I.) Random TCD screening at four centers suggests prevalence rates of the high risk state comparable to those reported by this investigative team in the pilot study on which trial assumptions are based. This trial is significant because it will determine if transfusion is effective in the primary prevention of stroke in Hb SS children at risk for stroke. Primary prevention is preferable to secondary prevention because it intervenes before the occurrence of brain injury. Secondary aims may further understanding of the effects of transfusion on the brain and guide future research into the cause(s) of cerebrovascular involvement in Hb SS.